| description abstract | Construction accidents can negatively affect the performance of a project. Evaluating the financial consequences of these accidents is an effective method for measuring their impacts on owners, contractors, and society. While several studies have estimated the cost of injuries among different trades, demographics, event types, injury sources, and nature of injuries, most of these studies presented descriptive statistics without analyzing any inferential statistics regarding the differences between average costs. To address this limitation, this study utilized robust measures of location (trimmed mean) and scale (Winsorized variance) in a three-way factorial ANOVA design to examine the effect of variables on the cost of injuries among electrical contractors. To create a reliable accident database of electrical contractor injuries, the team aggregated 388 nonfatal accidents collected from the 2016 Occupational Safety and Health Administration (OSHA) database. To determine the effectiveness of inferential statistical methods as applied to injury-cost analysis, null hypotheses were tested using three robust methods: (1) the Welch-type procedure; (2) an extension of Yuen’s method; and (3) percentile bootstrapping. The results of the study confirm that robust hypothesis testing approaches can be successfully implemented on safety data even when the assumptions of conventional test statistics are violated. In terms of the electrical case study, the outcomes indicate that various event types and project end-uses can change the cost of injuries among electrical contractors. Specifically, the findings revealed that caught in/between and exposure to electricity accidents can, on average, lead to higher injury costs than fall to lower levels and struck-by objects/equipment accidents. In terms of project’s end-use, construction accidents, on average, resulted in costlier injuries in nonbuilding projects than in buildings. The methodological results of this study can help contractors better quantify the risks of a construction project/task by estimating the severity of potential accidents in monetary values. Furthermore, this study contributes to the current body of safety knowledge by assessing alternative methods of hypothesis testing that do not require specific assumptions. | |
